tele9752wikiaorg-20200213-history
Xx7T
Background: Know: Recognize: ping 'Performance measurement tools' To measure the performance of network, we mainly focus on two indicators which are response and throughput. 1. 'Usually we use three commands “ping”, “hping” and “pathping” to check the response from one end. The response is also known as delay which means the time that data like a packet or a message transmitted from one end to another end in network. Network delay includes transmit delay, propagation delay and processing delay. Transmit delay indicates that the time of host or router transmitting a frame. Propagation delay indicates that the time cost by the frame transmitted in the medium in a certain distance. And the processing delay indicates that when host or router receives the frame, they need to analysis the header and the data, the whole processing time introduces delay. Then the total delay is the summation of transmit delay, propagation delay and processing delay. '- pathping: pathping is used in windows and includes ping and traceroute. It can be used to allocate the router or subnet with problem in a network, which is based on the packet loss of a certain router or link node. The format of pathping is: pathping -nMaximumHopsHostListPeriod NumQueries [-w Timeout IPAddress IPv4 IPv6TargetName. For example, pathping www.google.com: Tracing route to www.google.com 74.125.237.116 over a maximum of 30 hops: 0 Y-HP.wireless.unse.edu.au 149.171.228.202 1 vlan73.unswer1.gw.unsw.edu.au 149.171.182.66 2 ombcr1-te-1-7.gw.unsw.edu.au 149.171.255.105 3 unswbr1-te-7-2.gw.unsw.edu.au 149.171.255.105 4 gigabitethernet1.er1.unsw.cpe.aarnet.net.au 202.158.202.233 5 ge-5-1-0.bb1.a.syd.aarnet.net.au 202.158.205.161 6 tengigabitethernet2-1.pe1.a.syd.aarnet.net.au 202.158.202.19 7 72.14.197.241 8 66.249.95.224 9 72.14.237.135 10 syd01s12-in-f20.1e100.net 74.125.237.116 Computing statistics for 250 seconds... Hop RTT Lost/Sent = Pct Lost/Sent = Pct Address 0 Y-HP.wireless.unse.edu.au 149.171.228.202 0/ 100 = 0% | 1 3ms 0/ 100 = 0% 0/ 100 = 0% vlan73.unswer1.gw.unsw.edu.au 149.171.182.66 0/ 100 = 0% | 2 3ms 0/ 100 = 1% 0/ 100 = 0% ombcr1-te-1-7.gw.unsw.edu.au 149.171.255.105 0/ 100 = 0% | 3 3ms 0/ 100 = 0% 0/ 100 = 0% unswbr1-te-7-2.gw.unsw.edu.au 149.171.255.105 0/ 100 = 0% | 4 5ms 0/ 100 = 0% 0/ 100 = 0% gigabitethernet1.er1.unsw.cpe.aarnet.net.au 202.158.202.233 0/ 100 = 0% | 5 5ms 0/ 100 = 0% 0/ 100 = 0% ge-5-1-0.bb1.a.syd.aarnet.net.au 202.158.205.161 0/ 100 = 0% | 6 5ms 0/ 100 = 0% 0/ 100 = 0% tengigabitethernet2-1.pe1.a.syd.aarnet.net.au 202.158.202.19 0/ 100 = 0% | 7 5ms 100/ 100 =100% 100/ 100 =100% 72.14.197.241 0/ 100 = 0% | 8 7ms 0/ 100 = 0% 0/ 100 = 0% 66.249.95.224 0/ 100 = 0% | 9 --- 100/ 100 = 100% 100/ 100 = 100% 72.14.237.135 0/ 100 = 0% | 10 5ms 0/ 100 = 0% 0/ 100 = 0% syd01s12-in-f20.1e100.net 74.125.237.116 Trace complete. 2. 'For testing the throughput between tow ends we use “netperf” and “ttcp” commands. In network, throughput means the average rate of a message transmitted in a communication channel without any frame loss. And throughput is usually measured in bits per second or packet per second. In other words, it means how many bits can be transmitted in one second in a certain network channel. '- netperf: netperf is mainly used for TCP and UDP network performance testing with bulk data transfer mode and request/response mode. Netper test results indicate that how fast a host can transfer data to another host and how fast the host can receive the data. Netperf has client/server work mode, which means netserver is the server end used to monitor client connection and netperf is the client end used to test network. At first, a connection should be established between client and server to test the configuration and results. Then establish another connection to test the throughput of the network. For TCP network performance, netperf can test single TCP connection with bulk data transmission, single TCP connection with client/server transaction and multiple TCP connections. For UDP network performance, netperf can test one-way bulk transmission from client to server and request/response transaction. The format of netperf is: netper options – options. '-ttcp:' ttcp is a useful tool for testing network throughput. First we need to establish a connection between two ends. Then after configuration we can get the throughput results by sending TCP messages from transmitter to receiver in the network. The format of ttcp is: ttcp -t -u -s port buflen size numbufs align offset format -D -v host [ ttcp-r -u -s port buflen size align offset format -B -T -v >out Category:All